Detailed morphological structure and phylogenetic relationships of Degeeriella punctifer (Phthiraptera: Philopteridae), a parasite of the bearded vulture Gypaetus barbatus (Accipitriformes: Accipitridae)

Habitat loss is one of the main threats to species survival and, in the case of parasites, it is their hosts that provide their habitat. Therefore, extinction even at local scale of host taxa also implies the extinction of their parasites in a process known as co-extinction. This is the case of the bearded vulture (Gypaetus barbatus), which almost became extinct at the beginning of the twentieth century. After several attempts, this species was successfully reintroduced into the Alps at the end of the twentieth century. We collected 25 lice specimens from an electrocuted bearded vulture from Susa (Italian Alps) that were morphologically identified as Degeeriella punctifer. Six individuals were studied by scanning electron microscopy, with particular emphasis on their cephalic sensorial structures, while four further specimens were characterized at molecular level by amplifying partial regions of the 12SrRNA, COX1 and elongation factor 1 alpha (EF-1) genes. From a morphological perspective, the number, type and arrangement of the sensillae on the two distal antennal segments is quite similar to that of other species of the family Philopteridae (Phthiraptera: Ischnocera). The mandibles and tarsal claws allow lice to cling firmly to their host’s feathers. Phylogenetic analyses help unravel the paraphyletic nature of the genus Degeeriella and demonstrate the clear differentiation between lice parasitizing Accipitriformes and Falconiformes, as well as the close relationship between D. punctifer, D. fulva, D. nisus and Capraiella sp. that, along with other genera, parasitize rollers (Aves: Coraciiformes).


Morphological analyses.
Eight individuals (four males, three females and one nymph) were mounted in Canada balsam following 26 for observation using light microscopy. The remaining specimens were temporarily mounted in water to obtain measurements. Measurements (in µm) of the length and width of head, thorax, abdomen and parameres of males from each specimen were taken using a Nikon Eclipse 80i optical microscope (Nikon, Tokyo, Japan) equipped with a differential optical camera with optical inference contrast (DIC), and a Nikon Digital Sight DS-U1 digital camera. Cephalic (head width/head length) and body (maximum width/total length) indices were also obtained for each specimen. A further six specimens (three males, two females and one nymph) were selected for observation using SEM. To eliminate particles adhered to the cuticle, specimens were placed in a solution of 70% ethanol and 30% ether for three days 27 . Subsequently, specimens were submitted to an ultrasound session lasting 20 min with a medium emission frequency at 30% intensity. Individuals were cleaned with distilled water and dehydrated with a series of increasing concentrations of ethanol, finishing in acetone. A critical point was obtained before gold coating. Observations were made using a Zeiss Merlin scanning electron microscope (Carl Zeiss, Germany).
DNA extraction and amplification. The genomic DNA of four lice was extracted using the DNeasy Blood and Tissue extraction kit (QIAGEN, Hilden, Germany) following the manufacturer's instructions in all except for the initial incubation time, which was set to 48 h, and the final elution volume, established as 100 μL. The mitochondrial cytochrome oxidase I (cox1), mitochondrial small subunit ribosomal RNA (12S rRNA gene), and the nuclear elongation factor 1 alpha (EF-1) were amplified (Table 2) using the following set of primers: (a) L6625 (5′-CCG GAT CCT TYT GRT TYT TYGGNCAYCC-3′) and H7005 (5′-CCG GAT CCACNACR TAR TANGTR TCR TG-3′) for cox1 28  www.nature.com/scientificreports/ AGC GAC GGG CGA TGT GT-3′) for 12SrRNA 29 ; and (c) EF1-For3 (5′-GGNGAC AAY GTT GGY TTC AAC G-3′) and Cho10 (5′-ACRGCVACKGTYTGHCKCAT GTC -3′) for EF-1 30 . The final volume for each PCR was 20 µL, including 2 µL of DNA, 4.8 µL of water, 1.6 µL of each primer, and 10 µL of DNA Polymerase MyFi (BioLine, Meridian Life Science Inc., Tauton, USA). In all cases, an additional sample without DNA was included as a negative control. The thermocycling profile was performed under the conditions described by Smith et al. 31 . PCR products were observed on a 1% agarose gel and, in cases of poor amplification, a second PCR was performed with a greater number of cycles and greater alignment time. The amplified material was then purified with the Nucleospin PCR and Gel Purification Clean-up kit (Machery-Nagel, Düren, Germany) and sent to Macrogen (Spain) for sequencing. Sequences were trimmed for low-quality reads and assembled in Geneious Prime 2019.2.1 (https:// www. genei ous. com). To confirm the specific identity, obtained sequences were compared with the NCBI database using BLAST (Basic Local Alignment Search Tool) 32 .
Phylogenetic analyses. The obtained sequences were aligned independently for each gene using the online version of Mafft (https:// mafft. cbrc. jp/ align ment/ server/) and comparing with other Philopteridae sequences retrieved from GenBank ( Table 1). The nucleotide alignment of the protein-coding gene (cox1) was edited manually, in frame, using Geneious. Four datasets were analysed: (i-iii) each gene independently and (iv) concatenated sequences for all three genes. Evolutionary models for each gene were chosen using the jModel-Test2 program 33 and phylogenetic inferences were obtained using Bayesian Inference (BI) and Maximum Likelihood (ML) analyses. BI was performed using MrBayes 34 and run for 1,000,000 generations; burn-in was set at the point at which the average standard deviation split frequencies was < 0.01. ML analysis was obtained using GARLI (Genetic Algorithm for Rapid Likelihood Inference 35 and support values for each node were obtained after 100 replications.

Results
Morphological analyses. Lice were morphologically identified as Degeeriella punctifer (Gervais, 1758) (Ischnocera: Philopteridae). This species is included in the phlyctopygus group 9 , which is characterized by a distinctive type of genitalia with penial sclerite (Fig. 1) and, normally, more than four sternocentral silks in segments III and IV. The following description of adult males and females, and measurements of nymphs, is based on the examination of specimens with optical and scanning electron microscopy. Males (n = 9). Head rounded with the marginal carina ventrally thinned anteriorly, with a narrow hyaline margin and a postantenal suture present (Figs. 1(2)-2(3)). The cephalic index is > 0.94 (Table 2). Mandibles are large and strongly sclerotized while the labial palps are short and unisegmented, with six basiconic setae at the apex ( Fig. 2(4)). The coni, which are lateral extensions of the head, lie just above the articulations of the antennae; the small simple eyes are elongated ( Fig. 2(5)). The antennae are composed of five segments, with a distinct arrangement of pores and placoid sensillae on the inner side of the final two segments and basiconic sensillae at the apex ( Fig. 2(6-9)). The sternal thoracic plate is subtriangular, with two anterior setae and four setae on the posterior margin ( Fig. 3(10)). The legs are robust and have three tibial processes, with two tarsal claws ( Fig. 3(11)). The spiracles are located on the lateral side of terguites II-VII ( Fig. 3(12,13)). Tergum II has an unscletorized central area and tergum III medial narrowing; the remaining terga are elongated and cover the entire width of the abdomen 9 , which give this species a distinctive dorsal appearance. Four setae are inserted into the genital plate, as described by 9 .
Females (n = 3). Their appearance is similar to that of males, except for body size as the females are larger than the males, above all in the abdomen (Fig. 3(14,15); Table 2). The terga of segments IX-XI and genital region are as described by Clay (1958). The chaetotaxy of both sexes coincides with 9 Clay's description (1958) for D. punctifer. The terminal part of the abdomen at ventral level in females is quite distinct from that of the males, and has a longitudinal genital opening ( Fig. 3(16,17)).
Nymphs (n = 7): specimens were smaller in length than adults ( Table 2) and less chitinized, which gives them a more transparent appearance.
Molecular and phylogenetic analyses. Four sequences of D. punctifer were obtained for the 12S rRNA and cox1 genes, with lengths varying between 564 and 596 bp, and between 436 and 437 bp, respectively. Additionally, we obtained two sequences for the EF-1 gene of D. punctifer with 360 bp each ( Table 1). The nucleotide similarity of the obtained sequences reached 99.5%, 98.6% and 90.4% for the 12S rRNA, cox1 and EF-1 sequences, respectively. Datasets 1-3 for single gene analyses included 12, 20 and 14 sequences for the 12S rRNA, cox1 and EF-1 genes, respectively, whereas the concatenated analysis for the three genes included a total of 21 sequences (Table 1). All datasets rendered the Degeeriella complex as paraphyletic (Fig. 5). All newly obtained sequences for D. punctifer were grouped in a monophyletic clade, except when the EF-1 gene was analysed independently (Dataset 3) (Fig. 4c). A clade formed by D. fulva, Capraiella sp. and D. nisus was consistently placed as sister to D. punctifer when the three concatenated genes were analysed, although it was only well supported by Bayesian posterior probabilities but not by ML bootstrap values (MB pp = 0.9; ML bootstrap = 42%) (Fig. 5). A similar relationship between D. punctifer, D. fulva and Capraiella sp. was observed when the 12S rRNA and the EF-1 genes were analysed independently, but not for the cox 1 gene from which the phylogenetic relationship could not be clearly identified (Fig. 4). Of the Degeeriella species occurring on Accipitriformes, D. regalis occupies an early-diverging position (Figs. 4b, 5). The analysed Degeeriella species occurring on Falconiformes (i.e. D. rufa and D. carrutthi) were consistently placed as sisters to Picicola capitatus (Fig. 4c). The other Picicola spp. analysed in this study

Discussion
This study reports for the first time the occurrence of D. punctifer in the Italian Alps having been reported previously from Afghanistan and Sikkim 9 , Lesotho 37 , Spain 10 and the Indian Himalayas 38 . The size of our specimens falls into the size range reported by Clay 9 for this lice species. In lice, the antennae are the main peripheral sensory organs 20 . The structure, morphology and function of antennae in Phthirapteran suborders and families are very similar in all species and follow a common pattern 21,39 . In fact, the shape, number and distribution pattern of sensillae at the distal tip of the antennae of D. punctifer correspond to those described for D. fulva, D. regalis, Craspedorrhynchus platystomus, Bueelia spp. and Upupicola upupae [40][41][42][43] .
One remarkable characteristic of the specimens of D. punctifer analysed in this study was the sturdy composition of the mandibles and tarsal claws (Figs. 2(4), 3(11)). Ischnoceran lice use their mandibles and claws, as observed in D. punctifer, for attaching themselves firmly to host feathers or to phoretic flies (e.g. hippoboscids) but compared to Amblyceran lice move around little when not on their hosts 18 .
The immature specimens studied were quite similar to others in terms of both morphology and size, and were classified as third-instar nymphs. Nevertheless, the first-and second-instar nymphs of D. punctifer are still unknown, so the morphological description of all developmental stages of this species remains incomplete 44 .
We provide here for the first time the molecular sequences of partial cox1, 12SrRNA and EF-1 of D. punctifer. From the phylogenetic relationships analysed in this study, the sequences of each of the genes obtained were grouped together in a monophyletic clade, the exception being the EF-1 tree (Fig. 4c). Even though the two EF-1 sequences of D. punctifer were the same length (360 bp), there were many more ambiguities regarding specific nucleotide positions in one of the sequences, indicated by the low percentage of similarity between them (90.1%). Nevertheless, the monophyletic and well-supported clade (Bayesian pp = 1.0 and ML bootstrap = 100%) formed by the four sequences obtained for both mitochondrial genes (cox1 and 12SrRNA) is evidence that the four specimens used for the molecular analyses correspond to the same specific entity. Furthermore, a monophyletic clade formed by D. punctifer, D. fulva and Capraiella sp. was consistent when the EF-1 gene was analysed independently and when the three genes were concatenated (Fig. 5).
In general, D. punctifer was found to be closely related to a clade consisting of D. fulva and Capraiella sp. + D. nisus (Fig. 5), all of which are parasites of Accipitriformes. The distant phylogenetic position of D. regalis, also present on Accipitriformes, and D. rufa and D. carruthi on Falconiformes, demonstrates that the genera Degeeriella is paraphyletic, as suggested by previous molecular studies. Johnson et al. 12 used a more extensive phylogenetic context, including other species from the Degeeriella complex but only cox1 and EF-1 sequences, to show a sister relationship between D. carruthi and Picicola spp. from African woodpeckers, and D. fulva and Capraiella sp. from rollers (Coraciidae). In addition, Catanach and Johnson 13 also extensively studied the phylogenetic relationships of Degeeriella spp. and show it to be a paraphyletic genus, with a significant geographical structure for a parasite species that broadly reflects the higher taxonomic level distribution (i.e. Order) of its hosts 13 . These studies, together with the phylogenetic context for D. punctifer provided in this study, suggest the need for a re-organization of the taxonomy of Degeeriella, Capraiella and Piciola species as the current taxonomy does not reflect their evolutionary relationships 12 .
Some but not all lice genera show strong evidence of cospeciation and this may be due-at least in part-to differences in their dispersive abilities 45 . This needs be tested within the genus Degeeriella, which includes species Table 2. Biometrical data obtained from the Degeeriella punctifer specimens analysed in this study. HL: head length; HW: head width; CeI: cephalic index; TL: thoracic length; TW: thoracic width; AL: abdominal length; AW: abdominal width; ToL: total length; CoI: corporal index; PaL: parameres length. Measurements are shown in micrometers and expressed as the mean ( X) ± standard deviation ( SD ), followed by the minimum (Min) and maximum (Max) values, and the confidence interval (CI) at 95%. n: number of specimens analysed. NA: not applicable.

Males (n = 9)
Females (n = 3) Nymphs (n = 7)  www.nature.com/scientificreports/ like D. fulva and D. rufa that have a wide range of host species. Nevertheless, we lack molecular data for > 60% of the Degeeriella taxa. Therefore, further information is needed if we are to perform a reconciliation analysis of the phylogenies of Degeeriella species and their respective hosts to characterize cospeciation, host switching, extinction or other macroevolutionary events that may affect these species 25,[45][46][47][48][49] .

X ± SD
Of the four lice species parasitizing the bearded vulture, we only found specimens of D. punctifer. Ischnoceran lice have a greater capacity to remain on host feathers than Amblyceran lice 18 , although in our case no Falcolipeurus quadripustulatus specimens were found. The reintroduced bearded vultures were bred and kept in captivity 50 and so it seems that these conditions did not negatively influence the survival of ectoparasitic lice on these captive birds 51 . Further studies are still required to detect whether or not the same situation occurs with the other lice species that parasitize the bearded vulture. Finally, if the term co-extinction was coined to describe the extinction of a host and its ectoparasitic lice 2,3 , then we believe it to be logical to talk about the 'co-reintroduction' or 'co-recovery' of a bird host and, at least, one of its ectoparasitic lice species.